Researchers used Google Street View to identify hundreds of elements of the built environment, such as buildings, green spaces, sidewalks, and roads, and how these elements relate to each other and the coronary arteries of the people living in these areas. We studied how it affects diseases.

Their findings are: european heart journal Today (Thursday), researchers showed that these factors can predict 63% of the variation in coronary heart disease risk by region.

Coronary heart disease, in which a buildup of fatty substances in the coronary arteries cuts off the blood supply to the heart, is one of the most common forms of cardiovascular disease.

Researchers believe that using Google Street View can help provide an overview of physical environmental risk factors in the built and natural environments, helping to not only understand the risk factors in these environments, but ultimately and could help build or adapt cities to be healthier. A place to live.

The study was led by Professors Sadeer Alkindi and Sanjay Rajagopalan of the Harrington Heart and Vascular Institute and Case Western Reserve University Hospital in Ohio, USA, and Dr. Zhuo Chen, a postdoctoral fellow in Professor Rajagopalan’s lab. .

We have always been interested in how both the built and natural environments influence cardiovascular disease. Here in America, zip codes are said to be a better predictor of heart disease than any measure of personal health. However, investigating how the environment affects large populations in multiple cities is not an easy task. Therefore, we used a machine vision-based approach to assess the association between the built environment and the prevalence of coronary heart disease in US cities. ”

Professor Sanjay Rajagopalan, Harrington Heart and Vascular Institute and Case Western Reserve University Hospital, Ohio, USA

The study included more than 500,000 Google Street View images of Detroit, Michigan. Kansas City, Missouri. Cleveland, Ohio. Brownsville, Texas. Fremont, California. Bellevue, Washington. and Denver, Colorado. The researchers also collected data on the prevalence of coronary heart disease based on “census tracts.” These are areas smaller than a U.S. zip code and home to an average of 4,000 people. The researchers used an approach called convolutional neural networks. A type of artificial intelligence that can recognize and analyze patterns in images to make predictions.

The study found that features of the built environment displayed in Google Street View images could predict 63% of the variation in coronary heart disease between these small areas in U.S. cities.

Professor Al Kindi further added, “Highlighting some of the key regions within the image provides a semi-qualitative interpretation of some of the thousands of features considered important in coronary heart disease. We also used an approach called attention mapping,” he added. For example, features such as green space and walkable roads were associated with lower risk, while other features such as unpaved roads were associated with higher risk, but these findings may Investigation is required.

“We have shown that a computer vision approach can be used to help identify environmental factors that influence cardiovascular risk. This could play a role in guiding heart-healthy urban planning. The fact that we can do this at scale is absolutely unique and important for urban planning.”

“Given future challenges such as climate change and demographic change, and with nearly 70% of the world’s population living in urban environments, we are using computer vision approaches that can provide minute details instantly to improve urban environments. There is a dire need to understand this at scale, at an unparalleled level,” Professor Rajagopalan said.

In an accompanying editorial, Dr. Rohan Khera of Yale University School of Medicine in the United States writes, “The association between residence and outcome often trumps that of known biological risk factors. , often summarized by the expression that a person’s postal code is: However, our ability to properly classify environmental risk factors relied on population surveys that track wealth, pollution, and local resources.

“The study by Chen et al. presents a novel and more comprehensive assessment of the built environment. The study creatively leverages Google’s panoramic Street View imagery to supplement the widely used mapping application. .”

“…an AI-powered approach to studying the link between the physical environment and cardiovascular health will help us understand that across our communities, measures of cardiovascular health are not strongly encoded solely in neighborhood appearance. It is important to use this information wisely to define strategic priorities to identify vulnerable communities and improve cardiovascular health in communities with the greatest need. It’s about redoubling our efforts to improve vascular health.”